I've been thinking about making my own radio. I understand fixed-frequency radios are the simplest and easiest, but a colleague is recommending a spread-spectrum radio instead. Now I hear that there are two flavors of spread spectrum: frequency hopping and direct sequence. Can you help me understand the differences?
Signed, Bob the (Radio) Builder
Wise Guy: Sure. Let's look at how these gizmos work.
Narrowband, fixed-frequency transmission is the easiest to implement, from both a transmission and reception standpoint. As the name implies, the carrier frequency is fixed. With these, the sensor information is "placed" onto the carrier signal using amplitude modulation (AM), frequency modulation (FM), or phase modulation (PM) techniques, or some combination thereof. The carrier frequency is modulated with the information resulting in a typical frequency spectral distribution as shown in Figure 1. The receiver must simply be tuned to the correct center frequency while the transmission is received and demodulated. Such a transmission is clearly not secure, because anyone in the reception area with a spectrum analyzer can see that a signal is being transmitted.
Figure 1. A fixed frequency, narrowband transmitter
Frequency-hopping spread spectrum (FHSS), a variant of narrowband transmission, is based on periodically changing the signal's carrier frequency. While at any given moment the transmitter is functioning as a simple narrowband transmitter, after a predetermined time, the frequency changes, causing the information-bearing signal to move. The carrier frequency moves in discrete steps, or hops, within a prescribed frequency band, and the net result is that over a period of time the transmission has been spread across a wide frequency band or spectrum.
The beauty of this system is that at any given moment there is the chance for a narrowband transmitter to jam (interfere with) the signal, but while the narrowband transmitter stays at a fixed frequency, the FHSS transmission moves to another carrier frequency where the narrowband jammer does not radiate.
The procedure for generating the varying carrier frequency is straightforward and illustrated in Figure 2. A voltage control oscillator (VCO) functions as a voltage-to-frequency (V2F) converter, so that voltage 1 yields frequency 1, voltage 2 yields frequency 2, and so on. A microprocessor (or dedicated analog/digital circuit) generates a pseudorandom number (PN) based on a generating poly-nomial (typically a Fibonacci number sequence). This digital number is converted into an analog voltage (via a simple digital-to-analog converter) and applied to the V2F. Therefore, depending on the PN value, the V2F generates a certain frequency that serves as the carrier for the data transmission. At predetermined times, the PN number changes, thereby changing the V2F output and hence the carrier frequency.
Figure 2. A general design for a frequency-hopping spread spectrum (FHSS)
A complicating factor is that the receiver must have prior knowledge of where, in frequency, the next transmission will occur. But this is easily overcome by programming the receiver with the starting values of the PN generator.
Direct-sequence spread spectrum (DSSS) transmission is a fundamentally different method used in situations prone to interference. It relies on a circuit to convolve the data stream with a spreading code. The method resists interference by mixing the data signal with a PN code, a noise-like sequence of bits or chips valued 0 and 1. The resulting signal bandwidth becomes much larger, hence the term "spread spectrum." The receiving system applies the same PN code to the received signal to extract/decode the information.
The essence of this method of spreading is that the RF carrier frequency remains fixed while the signal being transmitted has been spread. The transmitter does not operate at the bit rate of the digital information, but rather at the higher bit rate associated with the coded signal. This causes the side lobe energy (Figure 3) to increase as the size of the spreading PN code increases. As more bits are added to the spreading code, more energy is transferred into the side lobes, thereby reducing the amplitude of any individual lobe. Notice that with a "suitable" size spreading code, it is possible to drop the energy in any single side lobe to less than a spectrum analyzer's noise floor, thereby essentially becoming undetectable.
Figure 3. A general design for a direct-sequence spread spectrum (DSSS) radio